I read a scientific paper, which is accompanied by a code library published by the authors. I couldn't find anything about its license.
Could I just reuse the code?
Is there a default license for code libraries that accompany scientific research papers? or a standard way to find out the license model?
Any intellectual property is copyrighted, including code, even if it is released. The authors still hold the copyright unless they explicitly relinquish it or permit certain uses of it. Most commonly this is done with a License or an agreement with a publisher. A publisher may hold the copyright to a publication but this may not include code and data released to support the manuscript (which may still be held or licensed by the authors).
It is best to check with the authors if it is possible to contact them. Generally they should not have a problem with if you intend to use it for academic, educational or research purposes (rather than for commercialisation or profit). In fact, they should encourage you to use their code and techniques to cite their work. Unless you are sure it is okay with them, you should not use it without a license. However, there are cases where the authors themselves are bound by the conditions of software that they’ve used.
There is no standard universal license for academic software. It is the choice of the authors but it is common for software developed for research purposes to be released with an open-source or copy-left License. It is worth checking that any of the code used includes the use of modifications of other open sourced code that does have a license. Many of these licences (e.g., GPL, MIT, BSD, or CC-BY) permit reuse and modification of the code, provided that the original authors are acknowledged appropriately. If software with any of these licenses is used, any code derived from them (even if it is modified) must also comply with the license of the original and similarly be free to distribute, reuse, or modify. How exactly the code can be reused or modified depends on the exact conditions of each license.
In the copyright law that most countries sign onto (e.g., Berne Convention), copyright is assigned upon creation. In other words, if there is no explicit permission, you are not allowed to redistribute it. This is why licensing is so important.
However, if the code is publicly available and you reuse part(s) of it to create your own creative work, you are the copyright holder of that new piece of code. If the original copyright holder is of the opinion you violated their copyright (i.e., they think you copied too much of it directly and made too little additions/changes of your own), you have to substantiate your claim on how you deviated or added sufficiently from the original work to get copyright yourself.
In other words, take the risk or ask for explicit permission/a license to be added (e.g., MIT, GPLv3; see also here).
In Berne Convention countries (which comprise almost all of the world), the source code is copyrighted as soon as it's written. Without a license, you can't make a copy (or do other things that can vary by country). So, without a license, you can't legally use it. You might be able to argue Fair Use in the US, but don't count on it, and check with a lawyer first.
Using it as a basis for your own development is (at least in the US) creating a derived work, and that's still under the initial copyright, so you still can't use it. You can start with the algorithms (which should be explained in the text part of the paper) and write your own.
The only way you can legally touch it is with permission. If you can't find a license, you can't use it. You can write to the copyright holders and ask for permission. You can suggest that they put some sort of open source license on it, but they aren't required to.
Using it internally and not revealing what you're doing will normally keep you from being sued, but it's still unlawful. Publicly using it would be an openly unlawful act, and I wouldn't try it without consulting a lawyer.
